The US and China are locked in a race for global dominance in artificial intelligence, competing across advanced chips, cutting-edge models, data centers and talent.
The competition is emerging as a central dimension of the rivalry between Washington and Beijing.
On Sunday, US Treasury Secretary Scott Bessent proposed an AI incident-notification system during talks with Chinese Vice Premier He Lifeng in New York, ahead of President Donald Trump’s planned meeting with Chinese President Xi Jinping.
Bessent told reporters that the American and Chinese officials had discussed creating a new "notification mechanism" for AI incidents that could affect national security.
But the two countries have different strengths across the AI industry, with the US holding an advantage in some areas and China gaining ground in others.
Here are the five main fronts shaping the US-China AI race:
Advanced AI requires enormous computing power, making semiconductors a foundation of the race.
US companies such as Nvidia and AMD remain leading designers of AI accelerators, while Taiwan Semiconductor Manufacturing Co. (TSMC) produces many of the world’s most advanced AI chips.
The US also holds important positions in chip-design software and semiconductor manufacturing equipment, giving Washington substantial influence over China’s access to cutting-edge computing.
Moreover, since 2022, the US Commerce Department has imposed and expanded controls restricting Chinese access to advanced processors, high-bandwidth memory, chipmaking equipment and design software.
Beijing has responded by accelerating efforts to develop domestic alternatives.
Chinese companies including Huawei are developing AI processors and computing systems intended to reduce reliance on foreign technology, while the government is pushing to localize semiconductor equipment, software and other parts of the supply chain.
Huawei has acknowledged, however, that it cannot yet meet China’s domestic demand for AI computing equipment.
The restrictions have restricted China’s access to leading-edge technologies in the short term, while adding urgency to its drive for greater technological self-sufficiency.
Washington has greater leverage at the leading edge of the semiconductor supply chain, while China is using its manufacturing base, domestic market and state support to develop local alternatives.
For years, the US held a clear lead in frontier AI models, but Chinese developers have narrowed the gap.
Stanford University’s 2026 AI Index says US and Chinese models have traded places at the top of performance rankings several times since early 2025.
As of March 2026, the top US model led the top Chinese model by 2.7%, with the gap remaining in single digits during the preceding year.
Epoch AI, an independent research institute, estimates that Chinese models have trailed the US frontier by an average of seven months since 2023, with the gap ranging from four to 14 months.
The US also continues to attract substantially more private AI investment, while China relies more heavily on government-backed funds and state-directed financing.
The competition is therefore shifting beyond which country has the single most capable model. Cost, efficiency, reliability and the ability to deploy models at scale are increasingly important sources of leverage.
AI leadership also depends on the physical infrastructure and electricity needed to train and operate models.
The US has the larger AI computing and data-center base, while China is rapidly expanding the power and industrial infrastructure that could support future growth.
The International Energy Agency says the US accounted for 45% of global data-center electricity consumption in 2024, compared with 25% for China and 15% for Europe.
Global data-center electricity use is projected to more than double by 2030, driven largely by AI and other digital services.
The US is expected to account for the largest share of the increase, followed by China. Data centers are projected to generate nearly half of total US electricity-demand growth through 2030.
China, meanwhile, is rapidly expanding its broader electricity and industrial infrastructure, potentially increasing its ability to support AI deployment at scale.
The IEA estimates that Chinese data centers consumed more than 100 terawatt-hours of electricity in 2024 and says their consumption could double by 2027, although projections remain uncertain.
AI capabilities ultimately depend on researchers and engineers.
The US remains home to more AI talent than any other country, supported by leading universities, technology companies, venture capital and frontier AI laboratories.
Its ability to attract additional talent, however, has declined.
According to Stanford’s 2026 AI Index, the number of AI researchers and developers moving to the US has fallen by 89% since 2017, including an 80% decline in the latest year measured.
China’s strength lies in the scale of its research and patent activity.
Stanford says China leads in AI publication volume and citations, while the World Intellectual Property Organization says it remained the dominant origin of generative-AI patents in 2025.
WIPO data show that Chinese inventors produced more than 38,000 generative-AI inventions between 2014 and 2023, about six times the US total over the same period.
The US, however, retains an advantage in the concentration of institutions and researchers developing frontier systems, as well as in the impact of its most influential patents.
AI systems depend not only on computing power and talent but also on the data used to train and improve them.
The US has a large digital economy and globally active technology platforms, while China combines extensive consumer data with industrial data generated by factories, vehicles, robots and other connected systems.
Such physical-world data can be used to develop AI applications for manufacturing, robotics and autonomous vehicles.
According to the International Federation of Robotics, China installed about 295,000 industrial robots in 2024, accounting for 54% of global installations and nearly nine times the US total.
The scale of China’s manufacturing and automation systems could provide a significant source of data for developing industrial AI.
Beijing is also seeking to make such information easier to combine and use. In 2026, China’s Industry Ministry launched a program aimed at organizing and standardizing high-quality industrial datasets for AI applications.
While the US benefits from its global technology platforms and broader AI ecosystem, China could gain an advantage from the scale of its manufacturing base and the industrial data it generates.
The five fronts show why the US-China AI competition cannot be reduced to a single question of which country is ahead.
Across chips, AI models, data centers and energy, talent, and data, the two countries hold different forms of leverage.
The US is stronger in advanced chip design, frontier AI development, existing computing infrastructure and the concentration of leading AI talent.
China has greater strengths in research scale, patent activity, manufacturing capacity, rapid infrastructure expansion and industrial data.
Different model rankings also show that assessments of the gap depend on how performance is measured. Stanford puts the difference between the leading US and Chinese models at 2.7%, while Epoch AI describes it as an average time lag of seven months.
The competition is therefore less about dominance on any single front than about which country can combine its advantages across these five areas and turn them into lasting leverage across the AI technology stack.
news_share_descriptionsubscription_contact
contact_the_ombudsman
is_there_an_error_in_this_story
